See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/220497374

A test of the efficacy of the MC Square device for improving verbal memory,
learning and attention

Article  in  International Journal of Learning Technology · January 2007

DOI: 10.1504/IJLT.2007.014844 · Source: DBLP

CITATIONS READS

4 127

4 authors, including:

Joseph I Tracy Michael Sperling

Thomas Jefferson University Thomas Jefferson University
64 PUBLICATIONS   1,051 CITATIONS    528 PUBLICATIONS   17,700 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Epilepsy View project

Network Reorganization in Temporal Lobe Epilepsy View project

All content following this page was uploaded by Michael Sperling on 14 March 2016.

The user has requested enhancement of the downloaded file.

 세계적 권위의 과학저널 IJLT 2007년판

- 엠씨스퀘어의 효능에 대한 연구논문 전격 수록 -

Int. J. Learning Technology, Vol. 3, No. 2, 2007 183
Copyright © 2007 Inderscience Enterprises Ltd.

Editorial
Lorna Uden
Staffordshire University, Stafford, UK
E-mail: L.uden@staffs.ac.uk
Biographical notes: Dr. Lorna Uden teaches computing in the Faculty of
Computing, Engineering and Technology at Staffordshire University. Her
research interests include Technology Learning, HCI, Activity Theory,
Knowledge management, Web Engineering, Multimedia, E-business and
Problem-Based Learning. She has published widely in conferences, journals
and chapters of books.
------------------------------------------------------------------------------------------------------------

There is a growing demand for learning tools that will aid and enhance performance
on standardised achievement or ability tests, employment/civil service tests, or for
those seeking admission to advanced schools. Although many of these cognitive
enhancement devices have positive anecdotal reports about them, there is little
evidence of empirical testing on their actual benefit. The authors of this paper set out
to evaluate the MC Square for its ability to improve key cognitive functions (verbal
learning, memory and attention) following substantive training and practice with the
device. A double-blind, placebo-controlled (sham device) and crossover design was
utilised with pre and post testing on the cognitive measures occurring during each
phase of the crossover.
The primary hypothesis was that after training with the MC Square there would be
improvement in verbal memory, associated learning, working memory and attention/
concentration. Results showed a statistically reliable improvement on the measure of
attention/concentration, and the digit-span forwards test, following MC Square
training.
According to these authors, the MC Square device provides modest enhancement in
the ability to focus, attend and report information over the short term.

------------------------------------------------------------------------------------------------------------
편집자 주:

최근에 각종 자격 시험, 취직시험, 입학시험 등에서 성적 향상 및 도움을 줄 수 있는 학습

기기에 대한 수요가 증가하고 있습니다. 비록 많은 인지 향상 기기들에 대한 긍정적인 자
료나 보도들은 많았지만, 실제 효과에 대한 검증적 실험은 거의 없었던 것이 현실입니다.
이번 논문의 발표자는 한 학습 기기를 가지고 충분한 사용 훈련을 시킨 후에 주요 인지 능
력인 언어학습, 기억력, 그리고 주의력과 같은 요소들을 향상시키는 효과를 가지고 있는지
를 평가하기 위해 엠씨스퀘어라는 기기를 평가해 보기로 실험에 착수하였습니다. 엄격한
double-blind, 플라시보 방법(실험자, 피실험자가 어떤 것이 진짜, 가짜 기기인지를 전혀 모
르는 상태의 실험) 및 크로스오버 디자인(사용자가 진짜 또는 가짜를 잘 모르는 상태에서
각각 일정 기간 교대로 사용)을 이용하여 각 크로스오버 기간 동안(진짜 또는 가짜 사용 기
간)에 일어나는 인지 능력의 측정을 사용 전후 실험을 통해 진행했습니다. 주요한 가설은
엠씨스퀘어를 사용하면 언어학습, 기억력, 주의력/집중력이 향상될 것이라는 것이었습니다.
엠씨스퀘어를 사용한 후에 주의력/집중력의 측정결과가 통계적으로 향상된 것을 발견할 수
있었습니다. 실험 발표자에 따르면, 엠씨스퀘어는 단기간에 정보를 집중, 주의 및 기억해
내는 능력을 향상시킨다는 것을 확인하였습니다.

------------------------------------------------------------------------------------------------------------

A test of the efficacy of the MC Square device for

improving verbal memory, learning and attention

Joseph I. Tracy*
Department of Neurology
Thomas Jefferson University
Suite 206, 900 Walnut Street
Philadelphia, PA 19107, USA
Fax: 215–503–2481
E-mail: Joseph.I.Tracy@jefferson.edu
*Corresponding author
Noman Ahmed, Waseem Khan and

Michael R. Sperling
E-mail: drnoman77@gmail.com
E-mail: drwaseem77@gmail.com
E-mail: Michael.Sperling@jefferson.edu
Abstract: Cognitive enhancement devices have been supported by positive anecdotal reports, but generally have not
undergone rigorous testing. In the following report we tested one such device, the MC Square, which uses Audio-

Visual Stimulation (AVS) (synchronised pulsed tones and flickering lights set at an alpha or theta frequency) to

entrain neural activity. Its effect on three key cognitive functions (verbal learning, memory, and attention) was tested
following a regimen of training with the device. A double blind, placebo controlled (sham device), and crossover

design was utilised with pre- and post-testing on the cognitive measures occurring during each phase of the crossover.

The primary hypothesis was that after training with the MC Square there would be improvement in verbal memory,

associative learning, working memory and attention/concentration. Results showed a statistically reliable

improvement on the measure of attention/concentration, the Digit Span Forwards test, following MC Square training.

The data suggest the MC Square device provides modest enhancement in the ability to focus, attend, and report

information over the short term.

Keywords: cognitive enhancement; neural training; memory; learning; attention; Audio-Visual Stimulation; AVS;
learning technology; MC square device.

Reference to this paper should be made as follows: Tracy, J.I., Ahmed, N.,Khan, W. and Sperling, M.R. (2007) ‘A
test of the efficacy of the MC Square device for improving verbal memory, learning and attention’, Int. J. Learning

Technology, Vol. 3, No. 2, pp.183–202.

Biographical notes: Joseph I. Tracy, PhD (ABPP-CN) is Clinical Associate Professor of Neurology and Radiology
at Thomas Jefferson University / Jefferson Medical College, Philadelphia. He is Director of the Neuropsychology

Division, and the Cognitive Neuroscience and Brain Imaging Laboratory.

184 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling

Noman Manzoor Ahmed, MD, is currently completing his Medical Residency at Wayne State University, Michigan.

Waseem Ahmad Khan, MD, is currently completing his Medical Residency at Nassau University Medical Center,

New York.
Michael R. Sperling, MD, is Baldwin Keyes Professor and Vice Chairman for Clinical Affairs at Thomas Jefferson

University/Jefferson Medical College, Philadelphia. He is Director of the Thomas Jefferson Comprehensive Epilepsy

Center, Department of Neurology.

1 Introduction
In modern society there is an increased focus on academic achievement. This competitive environment
has lead to a growing demand for learning tools that will aid and enhance performance on standardised
achievement or ability tests, employment/civil service tests, or for those seeking admission to advanced
schooling such as undergraduate or graduate school. In addition, cognitive enhancement tools are also
sought for rehabilitation after brain injury following stroke, head trauma, or other brain insults. Many of
these cognitive enhancement devices have positive anecdotal reports behind them, but have not
undergone rigorous testing to determine if they yield any actual cognitive benefit. In the following report
we provide the results for one such cognitive enhancement device, called the MC Square device. We
specifically evaluated the MC Square, for its ability to improve key cognitive functions (verbal learning,
memory, and attention) following substantive training and practice with the device.
The MC Square was developed by Daeyang E&C, Inc. of Seoul, Korea. This utilizes synchronised sound
and light to entrain brain waves to alpha and theta neural rhythms.
The device has gone through testing in laboratories in China, Korea, and Japan, but none of the reports
have been subject to peer review and thus will not be summarised or presented here. The device purports
to improve learning, memory, and attention and has sold over one million units in Korea, largely to
students. The device uses a series of flashing red lights in conjunction with pulsed tones and background
relaxing sounds (river gurgling, birds chirping) to achieve its effects. The lights are presented through an
eye goggle device that resembles a thick pair of eyeglasses. Light emitting diodes present red light that
appear as flickering dots which occur synchronously with pulsed tones at a rate and pattern that is thought
to induce alpha and theta brain wave activity. The technology differs from most available tools which
appear to focus on computer assisted learning or external memory aid devices such as the palm pilot,
calendars, and computer programmed memory reminders.
This technique of inducing alpha and theta waves of brain by Audio-visual Stimulation (AVS) is known
as brain wave entrainment or Audio-visual Entrainment (AVE). AVE has been demonstrated to cause
significant changes in EEG patterns and cerebral synchronisation. Scientific research examining the
effects of light and sound started in mid-1930s when scientists discovered that the electrical rhythm of
brain tended to adopt the frequency of light when this is used as external stimulation. In one of the earliest
reports, Adrian and Matthews (1934) confirmed that alpha rhythm can be driven above and below the
natural frequency by photic stimulation. Flickering light appears to A test of the efficacy of the MC Square
device 185 share some similarity in terms of frequency with brain waves in the alpha and theta range.
Manufacturers of light and sound devices have almost exclusively used red light-emitting diodes because
they are bright, inexpensive, and blood vessels in the eyelids pass red/orange light most efficiently.
Komatsu (1987) examined college students and found that red light produces optimal EEG driving it in
the 17–18 Hz band. Green increases
brain wave activity to 15 Hz, blue light enhances 10–13 Hz activity, and white light peaks at 18–19 Hz.
AVE has been associated with increases in cerebral blood flow (Fox and Raichle, 1985; Sappey-Marinier
et al., 1992) and this is thought to be one of the mechanism by which it entrains brain waves to the alpha
and theta state as measured by EEG. Fox and Raichle (1985) showed that photic stimulation at alpha and
low beta frequencies increased cerebral blood flow 20%–30% over baseline in the striate cortex.
Moreover, certain parameters of the EEG tend to correlate with cerebral perfusion at least in the
neocortex (Fried, 1993). Hypoperfusion will tend to be mirrored by the increase of theta band (4–8 Hz)
power in the EEG on the scalp surface in that location. Reductions in cerebral perfusion has been shown
to decrease with age and in elderly individuals generally, with some signs that the effect may be more
pronounced in the elderly who show cognitive deficits (Schreiter-Gasser et al., 1993). One well-
demonstrated effect of AVS is relaxation. This effect may arise from high sympathetic activation that
occurs during alpha state. EMG correlates of relaxation have been observed in individuals undergoing
AVS (Manns et al., 1981). Alpha and theta brain waves are considered optimal for learning and attention
and there have been attempts to induce these states to reduce memory problems and regain cognitive
function. Several studies have shown that there is a strong relationship between peak alpha rhythm and
mental performance (Jausovec, 1996). Klimesch et al. (1997) presented evidence that EEG oscillations in
the alpha and theta band reflect cognitive and memory performance in particular. A peak alpha rhythm of
less than 10 Hz is associated with poorer academic performance and an alpha rhythm frequency of more
than 10 Hz is associated with better performance (Jausovec, 1996). Budzynski and Tang (1998) collected
EEG in a sample of college students and subdivided alpha rhythms (9–13 Hz) into three categories (A1,
7–9; A2, 9–11; A3, 11–13) and examined whether the ratio between A3/A1 predicted academic
performance. A ratio value above 1.0 was associated with above average academic performance. They
also found that after 34 sessions of 14 Hz light stimulation the high-to-low alpha frequency ratio was
increased along with an increase in peak alpha frequency. In a later study by Budzynski et al. (1999)
again with college students, they found that following 30 sessions of repeated cycles of AVE at 22 Hz and
14 Hz in an alternating pattern, there was a significant increase in the mean A3/A1 ratio, alpha rhythm,
and academic performance. This positive ratio was also related to improved cognitive performance as
measured by a digit span task. Budzynski et al. (2002) used AVE (a Digital Audio-Visual Integration
Device, Paradise XL) to aid 31 elderly individuals who were experiencing cognitive problems. The AVE
session utilized random frequency stimulation from 9–22 Hz and an average of 33 treatment sessions took
place. The treatment was considered very cost effective because ten individuals could be treated at one
time. A computer-based Continuous Performance Test and the Microcog Test Battery were utilised to
assess cognitive change. The Microcog measures several domains of cognitive function including
attention, reasoning ability, memory, spatial ability, processing speed and accuracy, and cognitive
proficiency. Over 60% of participants showed improvement in at least some of the cognitive measures.
This AVE
186 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling
procedure has also been shown to improve cognitive functioning in certain clinical populations such as
dementia (Tan et al., 1997) and dyslexia (Magnan et al., 2004). Other studies with AVE devices have
suggested beneficial effects may be observed in behaviour and psychiatric symptoms such as depression
(Kumano et al., 1996; Rosenfeld, 1997), premenstrual syndrome (David, 1997) and attention deficit
disorder (Cohen and Douglas, 1972; Zentall and Zentall, 1976).
The current study investigated the cognitive efficacy of the MC Square device, to examine its affect on
major domains of cognitive functioning. Verbal material was chosen, since informal reports from users of
the device suggested that it produces gains in the acquisition and retention of verbal material such as
might be required when studying for an academic exam. As practice with the device was also reported to
be superior to single instance use, we implemented week long device use in testing it. A Sham device was
constructed to create a placebo arm of the study whereby participants were run through the identical
procedure without the key element of the MC Square device; that is the sham device used randomised not
synchronised light and sound.
Our hypotheses were as follows:
• After training with the MC Square Device there would be improvement in verbal memory, associative
learning, working memory and attention/concentration. The sham device will produce no such training
effect.
• On individual performance measures, pre- or post-training, the MC Square Device would be associated
with better performance than the sham device. The above cognitive tasks involve novel material that
requires effort, cognitive resources, and cognitive skill to complete successfully. In contrast, Vocabulary
items involving over-learned, highly familiar material that require no new learning and fewer cognitive
resources to complete successfully were used as a control task. These multiple-choice vocabulary items
called upon existing, readily available knowledge. Therefore, hypothesis three was that neither training on
the MC Square nor the Sham would produce improvement in Vocabulary.

2 Methods
Participants were recruited by advertisements at Thomas Jefferson University. All participants for this
study were screened for good general health and the absence of any neurological, psychiatric or medical
disorder. A questionnaire was constructed to eliminate individuals with photosensitive seizures, i.e.,
seizures in response to light
stimulation. A total of 120 normal, healthy adult subjects within the 18–45 age range were screened to
enrol 40 subjects. All subjects were native English speakers with at least an Average IQ (90 or greater)
based on the Shipley Hartford Institute for Living Scale. All participants were medical students, physical
therapy or PhD students, residents and nurses from Thomas Jefferson University. Ineligibility arose from
having IQ below 90, abnormal state or trait anxiety scores on the Spielberger Inventories, and prior
medical or psychiatric history with potential central nervous system impact (e.g., neurological or medical
condition with central nervous system impact, depression, anxiety, substance abuse, obsessive-
compulsive disorder, and migraines; 70 individuals).
Six individuals were dropped because of time and scheduling constraints. Three individuals were dropped
because of IQ less than 89, and one because of risk of A test of the efficacy of the MC Square device 187
photosensitive seizures. This produced a final enrolment sample of 40 subjects. One subject dropped out
in the middle of the study due to scheduling conflicts. This yielded a final analytic sample of 39.
Sample demographics can be seen in Table 1. The sample primarily was Caucasian though some mixed
ethnicity was present. The sample was well educated and of above average IQ. As the MC Square device
might induce relaxation, we sought to reduce the differential and beneficial effect this might have across
individuals by limiting our sample to individuals who showed clearly average range (i.e., low) levels of
state and trait anxiety as measured by the Spielberger Inventory (Spielberger et al., 1970).

Table 1 Demographic baseline screening data

2.1 Research design

The study utilised a double blind, placebo controlled, and crossover design. Within each element of the
crossover (MC Square device, Sham) participants underwent pre-testing on the cognitive measures,
training with the device, then post-testing on the same cognitive measures. Participants undertook
baseline testing (3 h) of the neurocognitive skills under investigation without the MC Square device. The
baseline testing included screening materials such as the The Shipley Hartford Institute of Living Scale
for assessment of IQ (Robert, 2001), and the Spielberger State and Trait Anxiety Scales for determining
baseline levels of anxiety. Other measures in this first session included the initial assessments of Memory
(Hopkins Verbal Learning Test) (Jason, 2001), Verbal Learning (Paired Associate subtests from the
Wechsler Memory Scale, WMS, with versions III, R, and the original WMS were used in order to obtain
five versions), Working Memory (Letter Number Sequencing subtest from the WMS-III with additional
versions constructed), (Wechsler, 1997) Attention (Digit Span subtest from the WMS-III with additional
versions constructed), and Vocabulary (multiple choice practice PSAT and SAT items, with the five
versions equated for difficulty by pre-testing with a separate sample of ten individuals. All five versions
were completed with the mean scores within two points of each other).
Participants underwent cognitive testing with initial use of the MC Square Device or Sham on Day 1.
Post-testing with the cognitive tests took place on Day 8. During the intervening period participants
practiced for 15–20 min each day at home with the MC Square device or Sham depending on the
experimental condition. A log book and sworn statement was used to attest to their practice with the
device. A second identical testing session was held on Day 15 (pre-test) and Day 22 (post-test) with the
device not utilized during the first session. Participants were randomly assigned in a counterbalanced
fashion
188 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling
to one of two experimental session orders, e.g., Sham (day 1 pre-test, day eight post-test) then the active
MC Square device (day 15 pre-test, day 22 post-test), or the opposite order – active MC Square then the
Sham device. During days nine though 14 participants were given a break and did not utilise the MC
Square device or Sham. All pre- and post-test sessions utilised the device and followed the sequence of
events depicted in Table 2. A post-doctoral research fellow administered the cognitive tests and remained
blind to participant assignment. In total, participants attended five sessions: baseline (3 h), session one
pre-test and post-test, and session two pre-test and post-test. The pre- and post-test sessions took
approximately two and a half hours. The blind was broken at the end of the last session for the last subject
after scoring and final entry of all data.

2.2 MC Square device characteristics

The MC Square device has different modes of operation, but only two modes were used in this study. The
P-1 mode is for concentration enhancement and set at an alpha frequency (8–12 Hz) throughout the study.
The P-2 mode is for inducing relaxation and is set at a combination of alpha and theta frequencies (4–12
Hz range), beginning with alpha, then theta (4–8 Hz), and then ending in alpha. The device works much
like a small hand held radio. The choice of modes was easily set by a dial on the surface of the device.
The device generated pulsed tones in the frequency range of 4–12 Hz (the range of theta and alpha waves),
which was always synchronised at the same frequency as the flickering light. In the background, nature
sounds were presented. The brightness of the flickering light, the volume of the tone pulses, and the
volume of the background nature sounds was adjustable. The rate of flickering of the light and the rate of
pulsing for the tones was not adjustable and determined by the P-1 and P-2 modes. There were four light
diodes per eye set at a flickering rate of 4–12 Hz with a wavelength of 70 to 370 nm and brightness of
400 to 1000 millicandels. The diodes emitted red light, and as noted earlier this is because blood vessels
in the eyelids pass red/orange light most efficiently.
The volume of the pulsed tones was allowed to be adjustable and customised to the subject so as to avoid
aversive loudness. The frequency set to alpha and theta, and the synchronisation of flickering lights and
tones were considered the major though not the sole ingredient of effective brain wave entrainment
through the MC Square device.
So that any observed effect could be reliably attributed to one variable, it was decided that the Sham
would be identical to the MC Square device with the exception of the frequency. The rate (frequency) of
the flickering light and the synchronised pulsed tone was randomised and never settled into an alpha,
theta or other range reliably for more than one second. Also, so as not to potentially eliminate a
characteristic that may contribute to the MC Square’s effectiveness, all other aspects associated with
typical use of the device were maintained (e.g., background audio track for relaxation).
Thus, the Sham device looked, felt, and operated identically to the active MC Square device with the
exception that the light and tone pulses were presented in a random, though synchronised, fashion and did
not utilise a wavelength entrainment algorithm.
During each pre- and post-test session the MC Square or Sham device was administered four times. The
order of events in the pre- and post-test sessions is depicted in Table 3. The device was kept in P-2 mode
for the first or initial use at each session then changed to P-1 mode for all subsequent uses. Subjects
utilised the P-1 mode during practice sessions at home.

A test of the efficacy of the MC Square device 189

Table 2 Experimental design and sequence of events
190 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling
Table 3 Pre- and post-training means and standard deviations of cognitive measures
during MC Square and Sham conditions

A
Table 3 Pre- and post-training means and standard deviations of cognitive measures
during MC Square and Sham conditions (continued)

Notes: Hopkins Verbal Learning Test-Total Recall (HVLT-TR).

Hopkins Verbal Learning Test-Delayed Recall (HVLT-DR).

Hopkins Verbal Learning Test-Percent Retention Score (HVLT-PR).

Hopkins Verbal Learning Test-Recognition Discrimination Index (HVLT-RDI).

Verbal paired Associates-Total Learning (VPA-TL).

Verbal paired Associates-Total Delayed Recall (VPA-TDR).

Verbal paired Associates-Learning Slope (VPA-LS).

Verbal paired Associates-Percent Retention (VPA-PR).

Verbal paired Associates-Delayed Recognition (VPA-DRecog.).

Letter Number Sequencing-(LN-Sequencing).

* Statistically significant pre/post difference after Bonferroni Correction.

The technical merits of the MC Square device are its portability, ease and comfort of use, relatively
inexpensive cost, safety, and potentially rapid effect on brain state.

2.3 Pre- and post-test measures

Four areas of neurocognitive functioning was assessed: Verbal Episodic Memory, Verbal Associative
Learning, Verbal Working Memory, and Attention/Concentration. Two of the tests utilised to assess these
domains had five versions available (Hopkins Verbal Learning Test, Paired Associates Learning Test).
Additional versions were constructed for the remaining tests (Vocabulary, Digit Span Forwards and
Backwards, and Letter Number Sequencing). A control task was developed, assessing vocabulary skills
through a multiple choice format. Pre-test and post-test sessions were identical with the exception
of the particular test version utilised. All tests with the exception of the Vocabulary test were well-
established, well-normed neuropsychological instruments. Administration and scoring procedures
followed the standardised procedure described in the test manuals.

192 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling

2.3.1 Verbal episodic memory

The Hopkins Verbal Learning Test (Jason, 2001) was used as a measure of verbal episodic memory. This
test consists of three learning trials composed of 12 individual words. A Delayed Recall of the words was
taken 20–25 after the third learning trial. A Delayed Recognition trial was then administered composed of
the 12 target words plus 12 non-target (non-heard) distracters. A Total Learning score (HVLT-TR) based
on the three learning trials, a Delayed Recall score (HVLT-DR) score, Delayed Recognition Index
(HVLT-RDI), and a Percent Retention score (HVLT-PR) were computed according to the manual and
utilised in the analyses.

2.3.2 Verbal associative learning

The Verbal Paired Associates subtest of the Wechsler Memory Scale-III (Wechsler, 1997) was used as a
measure of verbal learning. In this test there were a total of eight word pairs. After the word pairs were
read aloud, participants were given the first word in the pair and required to provide the second. The set
of eight word pairs was given and tested four times in each session. After 25–30 min a delayed recall test
was given where again only the first word of the pair was given and participants had to provide the
second word in the pair. A final delayed recognition phase was administered at each session.
Here, participants read the eight word pairs randomly interspersed among 16 non-target (non-heard) pairs
and participants had to identify the correct pair. Note, only non-semantically related word pairs were
utilised. By utilising past versions of the Wechsler Memory Scale (version III, Revised, and original, five
versions of this tests were created). A Total Learning score (VPA-TL) based on the four learning trials
was computed in addition to a Learning Slope score (VPA-LS), Total Delayed Recall score (VPA-TDR),
a Percent Retention score (VPA-PR), and a Total Delayed Recognition score (VPA-DRecog).

2.3.3 Verbal working memory

The Letter Number Sequencing subtest of the Wechsler Memory Scale-III (Wechsler, 1997) was utilised
as the measure of verbal working memory. Participants were given series of random letters and numbers.
The subject was required to repeat back the numbers in their cardinal (ascending) order and the letters in
alphabetical order. The string of letters and numbers became increasingly longer with each trial. The
subjects proceeded until they failed three times at a given letter/number span. A total of 21 trials were
administered, two at each span with a maximum span of eight items (four letters, four numbers).
Additional versions of this test were developed simply by utilizing strings of random numbers and letters
with no repetition of any string. The Total Score (LN-Sequencing) reflecting the number of correct trials
was utilised in the analyses.
The Digit Span Backwards subtest of the Wechsler Memory Scale-III was used as a second measure of
verbal working memory. The task is identical to the Digit Span Forward subtest described above with the
exception that the participant had to recite the digits backwards, opposite the heard order. Additional
versions of this test were developed simply by utilising strings of random numbers. The Total Digits
Backward score was in used in the analyses.
A test of the efficacy of the MC Square device 193

2.3.4 Attention/Concentration
The Digit Span (Forward) subtest of the Wechsler Memory Scale-III (Wechsler, 1997) was used as a
measure of auditory attention. The subject is required to repeat back a given string of digits, with each
series of digits becoming increasingly longer. The subjects proceed until they fail twice at a given digit
span. There were 16 trials, two at each span with a maximum span of nine digits. Additional versions of
this test were developed simply by utilising strings of random numbers. The total Digits Forward raw
score was used in the analysis.

2.3.5 Control task

A multiple choice Vocabulary test was used as a control task. It measures semantic knowledge. These
items were taken from preparatory books for standardised college entrance exams (PSAT, SAT). Target
items were presented with four choices and participants had to identify the synonym or word closest in
meaning to the target word. Pretesting was used to develop five equivalent 50-item versions. As noted,
improvement via the MC Square device was not expected for this measure.

2.4 Statistical analyses

Statistical analyses involved a repeated measures analysis of variance on the pre and post test scores of
the two sessions with two within-subject factors both two-level in nature: Session (First, Second),
Training Condition (pre-test, post-test). Order (MC Square condition then Sham, or vice versa) served as
a between subject factor. Additional
analyses of variance were run on pre- or post-test scores with Experimental Condition (MC Square,
Sham) and Session (First, Second) as between-subject factors. All analyses were subject to Type I error
correction using the Bonferroni method for 14 tests. An observed alpha of p < .001 was required to
maintain an effective alpha of p < .05.

3 Results
The sample mean scores and standard deviation for each measure are shown in Table 3.
We will first examine the effect of training with the MC Square device by focusing on the Repeated
Measures Analysis. Here, the key effect in the model testing the hypothesis of improved performance
following MC Square training involves the interaction between Session, Training Condition, and Order.
The statistical results for the Repeated Measures Analysis are shown in Table 4.
Of the 14 Repeated Measures Analyses of Variance the triple interaction of Session, Training Condition,
and Order was significant for Verbal Paired Associates Total Recall (F{1,37} = 6.975, p < .05), Digits
Forward (F{1,37} = 12.846, p < .01) and Backwards (F{1,37} = 6.104, p < .05), and Total Digit Span
(F{1,37} = 14.988, p < .01). In each instance the difference between the pre- and post-training scores was
greater in the MC Square condition compared to Sham with improved scores at the post-training session.
However, when Bonferroni correction for Type I Error was applied, only the effect for Digit Span
Forward (see Figure 1) and Total Digit Span (see Figure 2) scores remained significant. This training
effect for Digit Span Forward and Total Digit Span remained 194 J.I. Tracy, N. Ahmed, W. Khan and M.R.
Sperlin significant after accounting for baseline Digit Span performance. This was tested by re-running
the Repeated Measures Analysis of Variance and including the baseline Digit Span scores as a covariate.
A similar check of the finding was conducted using the Spielberger Trait and State Anxiety measures as
covariates (each run in separate models); again, the results (triple interaction) or Digit Span Forward and
Total Digit Span remained significant. On the Digit Span Forward test a total of 24 participants improved
at least a .5 standard deviation relative to their baseline. Seven subjects showed no change and eight
showed a decline. The average number of digits improved relative to baseline was .73.

Table 4 Results from repeated measures analysis of variance

Domain/Dependent variable with model effects F df P Value Verbal episodic memory
1 HVLT-Total Recall

Session 14.05 1,37 0.001

Session*Order 0.824 1,37 0.37

Training Condition 3.37 1,37 0.074

Training Condition*Order 0.025 1,37 0.876

Session* Training Condition 0.011 1,37 0.916

Session*Training Condition*Order 0.42 1,37 0.521

2 HVLT-Delayed Recall

Session 2.09 1,37 0.156

Session*Order 5.22 1,37 0.028

Training Condition 4.7 1,37 0.037

Training Condition*Order 0.648 1,37 0.426

Session* Training Condition 1.2 1,37 0.267

Session*Training Condition*Order 2.01 1,37 0.164

3 HVLT-Percent Retention

Session 0.425 1,37 0.518

Session*Order 0.379 1,37 0.542

Training Condition 1.337 1,37 0.255

Training Condition*Order 0.046 1,37 0.832

Session* Training Condition 0.036 1,37 0.85

Session*Training Condition*Order 0.262 1,37 0.612

4 HVLT-Recognition Discrimination Index

Session 1.46 1,37 0.234

Session*Order 0.802 1,37 0.376

Training Condition 3.04 1,37 0.089

Training Condition*Order 0.02 1,37 0.887

Session* Training Condition 1.02 1,37 0.317

Session*Training Condition*Order 2.16 1,37 0.15

A test of the efficacy of the MC Square device 195

Table 4 Results from repeated measures analysis of variance (continued)

Domain/Dependent variable with model effects F df P Value

Verbal associative learning

1 VPA-Total Learning

Session 10.925 1,37 0.002

Session*Order 10.064 1,37 0.003

Training Condition 7.274 1,37 0.01

Training Condition*Order 0.151 1,37 0.7

Session* Training Condition 8.98 1,37 0.005

Session*Training Condition*Order 6.975 1,37 0.012

2 VPA-Total Delayed Recall

Session 6.214 1,37 0.017

Session*Order 0.041 1,37 0.841

Training Condition 1.868 1,37 0.18

Training Condition*Order 0.119 1,37 0.732

Session* Training Condition 6.375 1,37 0.016

Session*Training Condition*Order 1.563 1,37 0.219

3 VPA-Learning Slope

Session 5.518 1,37 0.024

Session*Order 4.375 1,37 0.043

Training Condition 0.77 1,37 0.386

Training Condition*Order 2.852 1,37 0.1

Session* Training Condition 0.054 1,37 0.817

Session*Training Condition*Order 2.419 1,37 0.128

4 VPA-Percent Retention

Session 0.459 1,37 0.502

Session*Order 0.249 1,37 0.621

Training Condition 0.721 1,37 0.401

Training Condition*Order 1.147 1,37 0.291

Session* Training Condition 0.889 1,37 0.352

Session*Training Condition*Order 0.31 1,37 0.581

5 VPA-Delayed Recognition
Session 0.768 1,37 0.386

Session*Order 0.768 1,37 0.386

Training Condition 0.768 1,37 0.386

Training Condition*Order 0.768 1,37 0.386

Session* Training Condition 0.768 1,37 0.386

Session*Training Condition*Order 0.768 1,37 0.386

196 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling

Table 4 Results from repeated measures analysis of variance (continued)

Domain/Dependent variable with model effects F df P Value
Verbal working memory

1 LN-Sequencing

Session 6.653 1,37 0.014

Session*Order 17.823 1,37 0

Training Condition 2.428 1,37 0.128

Training Condition*Order 0.282 1,37 0.599

Session* Training Condition 1.192 1,37 0.282

Session*Training Condition*Order 3.208 1,37 0.081

2 Digit Span (Backward)

Session 5.93 1,37 0.2

Session*Order 2.52 1,37 0.121

Training Condition 5.822 1,37 0.021

Training Condition*Order 0.355 1,37 0.555

Session* Training Condition 0.004 1,37 0.949

Session*Training Condition*Order 6.104 1,37 0.018

Attention/Concentration

1 Digit Span (Forward)

Session 0.115 1,37 0.736

Session*Order 3.468 1,37 0.071

Training Condition 2.229 1,37 0.144

Training Condition*Order 1.365 1,37 0.25

Session* Training Condition 0.274 1,37 0.604

Session*Training Condition*Order 12.846 1,37 0.001

2 Total Digit Span

Session 3.481 1,37 0.07

Session*Order 5.161 1,37 0.029

Training Condition 6.673 1,37 0.014

Training Condition*Order 0.086 1,37 0.771

Session* Training Condition 0.226 1,37 0.609

Session*Training Condition*Order 14.988 1,37 0.001

Control Task

Vocabulary

Session 1.114 1,37 0.298

Session*Order 1.735 1,37 0.196

Training Condition 0.51 1,37 0.48

Training Condition*Order 2.101 1,37 0.156

Session*Training Condition 0.438 1,37 0.512

Session*Training Condition*Order 0.001 1,37 0.971

A test of the efficacy of the MC Square device 197

Figure 1 Mean raw scores on the Digit Span Forward test

Figure 2 Mean raw scores on the Total Digit Span measure

F{1,37}=14.98,P<.00

198 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling

The Analyses of Variance run on pre- or post-test scores with Experimental Condition (MC Square,
Sham) and Session (First, Second) as between-subject factors revealed a significant advantage for the MC
Square device at post-training for the Hopkins Verbal Learning Test Delayed Recall (Wechsler, 1997),
Letter Number Sequencing (Wechsler, 1997), Digit Span Forward (Wechsler, 1997), Backward and Total
Score (Wechsler, 1997). After accounting for Type I error through the Bonferroni correction (28 tests),
none remained significant. When we collapsed across the pre- and post-test scores and utilised their mean
score, an advantage for performance under the MC Square device was evident on the Letter Number
Sequencing task that was initially significant (p < .05), but this effect became non-significant under
Bonferroni Correction (14 tests).

4 Discussion
We conclude that there was a statistically reliable improvement on a measure of attention and
concentration, the Digit Span Forwards test, following MC Square training. There was improvement on a
measure of associative verbal learning and working memory in the initial analyses, but these findings did
not survive Bonferroni correction. As expected the MC Square device had no influence on our control
task involving vocabulary. The lack of an effect in most areas measured, including our control task,
provided assurance that the MC Square was not having a general effect on cognitive activity and that
when reliable change occurred it was a fairly specific effect. The training effect on Digit Span Forwards
held true even after accounting for the participant’s baseline level of Digit Span skill or their state and
trait level of anxiety. A total of 24 out of 39 subjects (61.5%) showed at least a half standard deviation
improvement (an increase of .73 digits, or approximately one digit) on the digit span task following
training with the device. This increase may be of practical benefit in terms over holding on to more
‘heard’ information over the short term.
In terms of performance on the individual tests, ignoring the training aspects of the study, participants
using the MC Square Device subjects showed generally better performance on a working memory
measure, the attention measure, and aspects of the associative learning test. However, here again these
effects can only be considered trends as they were not statistically reliable after accounting for potential
error rates (Type I error) that can occur when conducting multiple statistical tests. The large number of
tests conducted (14) certainly worked against obtaining statistically robust results that could remain
significant after the stringent Bonferroni test. A Type II error, particularly in the case of the Associative
Learning (Verbal Paired Associates, Total Recall) and Working Memory measures (Digits Backwards)
was certainly possible. Initial analyses suggested an effect that was not sustained with Bonferroni
correction. Note, in the repeated measures ANOVA examining the training effect for Verbal Paired
Associates Total Recall the power present was .73. The power estimate for Digit Span Backwards was .67.
Note, that the power for the training effect that did remain significant after Bonferroni correction, Digit
Span Forwards, was .94. Therefore, power was at adequate levels and the relatively small sample size did
not play as large a role in the loss of statistical significance as did the multiple tests and Bonferroni
criteria. A larger sample size may show an effect for verbal learning, and further study is warranted.

A test of the efficacy of the MC Square device 199

We observed improvement in response to an AVE device on the same task, the digit span task, as did
Budzynski and colleagues (1999). It is important to note that our effect was achieved with fewer training
sessions. Also, our finding of increased attentional skill is interesting in light of previous results that
suggest AVE devices can reduce inattention, impulsiveness, and reaction time in Attention Deficit
Disorder children (Cohen and Douglas, 1972; Joyce and Siever, 2000).
The exact neural mechanism behind audio-visual entrainment is not known. We did not collect EEG data
on brain wave activity on our participants and thus cannot confirm that entrainment actually occurred, or
that it involved a switch to alpha or theta rhythms.
Also, it is unclear whether the MC Square works through relaxation or a more direct impact on the brain
areas implementing specific cognitive functions such as attention.
Though baseline anxiety did not account for the effects on attention we observed here, we did not directly
measure anxiety or stress through techniques such as skin conductance while subjects were performing
the tests. This would allow determination of whether measures of physiologic arousal and relaxation
taken during cognitive performance are related to and account for the level of skill or accuracy that is
displayed.
It is difficult to know whether the improvement we found in attention would translate into improved test
performance in academic settings. The data do suggest the device provides some modest enhancement in
the ability to focus, attend, and report information over the short term. There were trends in the data
indicating the MC Square may have a direct effect on verbal learning, as indicated by the initially
significant finding for Verbal Paired Associates, yet the reliability of this effect is unclear. It certainly
points to a potential area of gain the device may confer, but replication studies with fewer, and more
selective set of measures focused on verbal learning will be needed.
Future studies should determine how underlying brain activation actually changes during use of the MC
Square device, particularly in a manner that correlates with improved attention, learning, and memory.
This could be accomplished through functional magnetic resonance imaging. Further, delineation and
confirmation of the device’s effect on EEG entrainment is also needed and is planned. Finally, as the
possible mechanism of action of the device in terms of influencing cognition may involve changes in
blood flow, a study verifying the MC Square device’s effect on blood flow is needed.
This could be accomplished by perfusion MR imaging. A technique that increases blood flow may at least
partially remediate certain cognitive deficits if applied regularly over a period of time, and could certainly
have application to clinical samples such as stroke and normal populations as the elderly where there is a
documented decrease in cerebral blood flow (Gur et al., 1987; Hagstadius and Risberg, 1989; Heiss et al.,
1992; Meyer et al., 1993; Nagahama et al., 1997). Improving cognitive performance is a goal of students
and others in the normal population who participate in college preparatory courses, occupational training
or lifelong learning programmes. This type of technology-enhanced learning could be more generally
applied to optimise worker efficiency in industry and organisations. As learning technologies to enhance
teaching shift to include more learnercentred approaches emphasising the situational contexts of the
learner, advances in optimising the biological state of the learner are needed (Uden, 2004). This is
particularly true as the teaching environment shifts to the medium of e-teaching where the learning is
highly individualised, and where the learner tends to work in isolation at their computer, at their own pace
(Bork, 2004). For such situations, where the learning environment is controlled by the individual, the MC
Square technology seems ideally suited. The search 200 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling
continues for tools that help improve brain function directly by effecting neural functioning, or indirectly
through the benefit of relaxation during periods of learning and test performance. Such practical
applications remain a goal of researchers in instructional technology, education, occupational training,
cognitive psychology, and neuroscience more broadly.

4.1 Definitions
Alpha Rhythm – The frequency for brain waves in the 8 to 12 Hz range. It is characteristic of a relaxed,
alert state of consciousness.
Theta Rhythm – The frequency for brain waves in the 4 to 8 Hz range. It is associated with drowsiness.
Cerebral Synchronisation – Cerebral synchronisation is a neurologic phenomenon that occurs when the
brain waves of both hemispheres are oscillating at the same rhythm or wavelength. In this state, the brain
works harmoniously and uses each hemisphere’s capacities equally. It can also refer to a state where the
separate cerebral hemispheres have symmetric brain waves.
Cerebral perfusion – The rate and volume of blood flow to the brain.
Peak Alpha Rhythm – A pattern of brain waves characterised by a maximum amplitude and frequency
that is consistently in the alpha range.
Photic stimulation – A type of treatment or exposure to the brain in which light stimuli is delivered to
eyes in a particular pattern or at a particular intensity with the goal of creating specific visual conditions.
The stimuli are often presented through goggles to ensure controlled administration.
Sympathetic Activation – Activation of the part of nervous system most sensitive to external stimuli and
the outside environment. This system is characterised by behavioural output involving either a ‘fight’ or
‘flight’ reaction in response to a perceived stimulus.
Alpha Entrainment – A technology whereby brain waves are triggered to consistently flow in the alpha
range. Entrainment refers to a specific technique where the triggering external stimulus such as a
flickering light or pulsing tone is set at a frequency close to alpha with the goal of training brain waves to
follow that same rhythm. Brain waves adopting an alpha pattern and inducing a relaxed/alert state of mind
because of this technique are said to have gone through alpha entrainment.

Acknowledgement
This project was funded by Daeyang, E&C, Inc., manufacturer of the McSquare device.
A test of the efficacy of the MC Square device 201
References
Adrian, E.D. and Matthews, B.H.C. (1934) ‘Potential changes from occipital lobes of man’, Brain,

Vol. 57, pp.355–385.

Bork, A. (2004) ‘A new lifelong learning system for the world’, Internal Journal of Continuing Engineering

Education and Lifelong Learning, Vol. 14, pp.46–57.

Budzynski, T., Budzynski, H., Sherlin, L. and Tang, H. (2002) ‘Short and long term effects of Audio Visual
Stimulation (AVS) on an Alzheimer’s Patient as documented by Quantitative Electroencephalography (QEEG) and

Low Resolution Electromagnetic brain Tomography (LORETA)’, Journal of Neurotherapy, Vol. 6, No. 1. Budzynski,

T., Jordy, J., Budzynski, H., Tang, H. and Claypoole, H. (1999) ‘Academic performance enhancement with photic

stimulation and EDR feedback’, Journal of Neurotherapy, Vol. 3, No. 3, pp.11–21.

Budzynski, T.H. and Tang, J. (1998) Biolight Effects on the EEG (SynchroMed Report), Seattle, WA.

Cohen, N. and Douglas, V. (1972) ‘Characteristics of the orienting response in hyperactive and normal children’,

Psychophysiology, Vol. 9, pp.238–245. David, N. (1997) ‘PMS, EEG, and photic stimulation’, J. of Neurotherapy,

Vol. 2, No. 2, pp.8–13.

Fox, P.T. and Raichle, M.E. (1985) ‘Stimulus rate determines regional brain blood flow in striate cortex’, Annals of

Neurology, Vol. 17, No. 3, pp.303–305.

Fried, R. (1993) What is Theta? Biofeedback & Self-Regulation, Vol. 18, pp.53–58.

Gur, R.C., Gur, R.E., Obrist, W., Skolnick, B. and Reivich, M. (1987) ‘Age and regional blood flow at rest and

during cognitive activity’, Archives of General Psychiatry, Vol. 44, pp.617–621.

Hagstadius, S. and Risberg, J. (1989) ‘Regional cerebral blood flow characteristics and variations with age in resting

normal subjects’, Brain and Cognition, Vol. 10, pp.28–43.

Heiss, W.D., Pawlik, G., Holthoff, V., Kessler, J. and Szelies, B. (1992) ‘PET correlates of normal and impaired

memory functions’, Cerebrovascular and Brain Metabolism Reviews, Vol. 4, pp.1–27.

Jason, B. (2001) Hopkins Verbal Learning Test-revised, New York: Psychological Assessment Resources. Jausovec,

N. (1996) ‘Differences in EEG alpha activity related to giftness’, Intelligence, Vol. 23, pp.159–173. Joyce, M. and

Siever, D. (2000) ‘Audio-visual entrainment program as a treatment for behavior disorders in a school setting’,

Journal of Neurotherapy, Vol. 4, No. 2, pp.9–25.

Klimesch, W., Doppelmayr, M., Pachinger, T. and Ripper, B. (1997) ‘Brain oscillations and human memory: EEG

correlates in the upper alpha and theta band’, Neuroscience Letters, Vol. 238, pp.9–12.
Komatsu, H. (1987) ‘Studies on the temporal frequency characteristics of vision by photic driving method: III

temporal frequency characteristics of color vision’, Tohoku Psychologica Folia, Vol. 96, Nos. 1–4, pp.1–12.

Kumano, H., Horie, H., Shidara, T., Kuboki, T., Suematsu, H. and Kindschi, C.L. (1996)

‘Treatment of depressive disorder patient with EEG-driven photic stimulation’, Biofeedback and Self-Regulation, Vol.

21, pp.323–334.

Magnan, A., Ecalle, J., Veuillet, E. and Collet, L. (2004) ‘The effects of an audio-visual training program in dyslexic

children’, Dyslexia, Vol. 10, No. 2, pp.131–140.

Manns, A., Miralles, R. and Adrian, H. (1981) ‘The application of audiostimulation and electromyographic

biofeedback to bruxism and myofascial pain-dysfunction syndrome’, Oral Surgery, Vol. 52, No. 3, pp.247–252.

Meyer, J.S., Terayama, Y. and Takashima, S. (1993) ‘Cerebral circulation in the elderly’, Cerebrovascular and Brain

Metabolism Reviews, Vol. 5, pp.122–146. 202 J.I. Tracy, N. Ahmed, W. Khan and M.R. Sperling Nagahama, Y.,
Fukuyama, H., Yamauchi, H., Katsumi, Y., Magata, Y., Shibasaki, H., et al. (1997)

‘Age-related changes in cerebral blood flow activation during a card sorting test’, Experimental Brain Research, Vol.
114, pp.571–577.

Robert, A.Z. (2001) Shipley Institute of Living Scale, LA.

Rosenfeld, P. (1997) ‘EEG biofeedback of frontal alpha asymmetry in affective disorders’, Biofeedback, Vol. 25, No.

1, pp.8–12.

Sappey-Marinier, D., Calabrese, G., Fein, G., Hugg, J., Biggins, C. and Weiner, M. (1992) ‘Effect of photic

stimulation on human visual cortex lactate and phosphates using 1H and 31P magnetic resonance spectroscopy’,

Journal of Cerebral Blood Flow and Metabolism, Vol. 12, No. 4, pp.584–592.

Schreiter-Gasser, U., Gasser, T. and Ziegler, P. (1993) ‘Quantitative EEG analysis in early onset Alzheimer’s disease:

a controlled study’, Electroencephalography and Clinica Neurophysiolology, Vol. 86, pp.15–22.

Spielberger, C.D., Gorsuch, R.L. and Lushene, R.E. (1970) STAI Manual for the State Trait Anxiety Inventory, Palo
Alto, CA: Consulting Psychologists Press.

Tan, G., Kelly, J. and Calhoun, W. (1997) ‘Brain stimulation to improve cognition and mood of geriatric patients

with dementia’, Paper Presented at the Association for Applied Psychophysiology and Biofeedback, Sandiogo, CA.

Uden, L. (2004) ‘Editorial: learning technology’, International Journal of Learning Technology, Vol. 1, pp.1–15.

Wechsler, D. (1997) Wechsler Memory Scale, 3rd ed., San Antonio, TX: Psychological Corporation.

Zentall, S. and Zentall, T. (1976) ‘Activity and task performance of hyperactive children as a function of

environmental stimulation’, Journal of Consulting and Clinical Psychology, Vol. 44, No. 5, pp.693–697.
 참조 1 : 엠씨스퀘어 IJLT과학 저널 논문 수록관련 보도기사

View publication stats